Understanding the cognitive processes and brain mechanisms with which words are stored and retrieved from the mental lexicon is critical to our understanding of how language is organized in the human brain. Fundamental aspects of this process, however, remain unclear. A major debate concerns whether complex words like 'teacup'are stored and retrieved in terms of smaller parts called morphemes (tea/cup), or whether all words are treated alike as whole-word chunks. As these views make fundamentally distinct predictions regarding the nature of word representations, investigating complex words becomes crucial. This project involves recording brain activity using high temporal resolution magnetoencephalography (MEG) during complex word processing, allowing real-time measurement of the stages of word recognition and their brain- level instantiation, prior to any overt judgment such as reaction time. Utilizing these time-sensitive brain-level measurements provides a new way forward in investigating whether the mental lexicon involves internally- structured representations, the time course and constraints on their decomposition, and what cortical mechanisms subserve these computations. Specifically, the timing and neural bases of complex word processing will be examined by taking advantage of the specific linguistic properties of compound words, allowing careful manipulation of variables that may constrain decomposition, in ways difficult or impossible when examining affixed (e.g. past-tense) words only. Establishing whether the brain initially processes both known complex words (e.g., 'teacup') and novel words which may have meaningful word parts (e.g. 'drugrack') in terms of those meaningful word parts will inform our understanding of whether the meaningful word part or the whole-word is the basic unit of representation in word recognition. Further, measuring brain responses to compound words with relatively transparent meaning relations (e.g. 'teacup') and those without (like 'honeymoon'), provides a direct test of whether, when, and with what underlying neural mechanisms words are recognized, and in particular whether meaningful word parts constitute a fundamental level of word representation independent of the levels of word form and word meaning. Developing a fuller neurocognitive model of word reading has direct health-related impact for addressing the range of language disorders involving word recognition and reading deficits, including dyslexia, Specific Language Impairment, aphasias, and Alzheimer's disease, increasing the potential for identifying new treatments for these deficits, many of which involve pervasive yet poorly understood problems with word parts. Further, the MEG assays developed may provide sensitive new tools for diagnosing and directly measuring brain-level effects of therapy in clinical intervention. This research project investigates the nature and cortical mechanisms underlying the reading of complex words. Better understanding the basic cognitive processes in word reading has direct impact for our understanding of language disorders, for which the reading of complex words presents fundamental, though still poorly understood challenges, including dyslexia, Specific Language Impairment, disorders associated with stroke and brain trauma, and diseases including Alzheimer's Disease and Parkinson's Disease. Both the findings and the innovative brain-level measurements of visual word recognition developed in this project have direct potential for application to the characterization, diagnosis, and remediation of these disorders.